Health resource utilization of labor induction versus expectant management

Abstract

Background:

Although induction of labor (IOL) of low-risk nulliparous women at 39 weeks reduces the risk of cesarean delivery compared with expectant management, concern regarding more frequent use of labor induction remains given that this intervention historically has been thought to incur greater resource utilization.

Objective:

To determine whether planned elective labor induction at 39 weeks among low-risk nulliparous women, compared with expectant management, was associated with differences in health care resource utilization from the time of randomization through 8 weeks postpartum.

Study Design:

This is a planned secondary analysis of a multi-center randomized trial in which low-risk nulliparous women were assigned to IOL at 39 weeks or expectant management. We assessed resource utilization post-randomization in three time periods: antepartum (AP), delivery admission, and discharge through 8 weeks postpartum (PP).

Results:

Of 6096 women with data available, those in the IOL group (n = 3059) were significantly less likely in the AP period after randomization to have at least one ambulatory visit for routine prenatal care (32.4% vs. 68.4%), unanticipated care (0.5% vs. 2.6%), or urgent care (16.2% vs. 44.3%), or at least one antepartum hospitalization (0.8% vs. 2.2%, p<0.001 for all). They also had fewer tests (e.g., sonograms, blood tests) and treatments (e.g., antibiotics, intravenous hydration) prior to delivery. During the delivery admission, women in the IOL group spent a longer time in labor and delivery (median: 0.83 vs. 0.57 days), but both women (p=0.002) and their neonates (p<0.001) had shorter postpartum stays. Women and neonates in both groups had similar frequencies of PP urgent care and hospital readmissions (p>0.05 for all).

Conclusions:

Women randomized to IOL had longer durations in labor and delivery, but significantly fewer AP visits, tests, and treatments, and shorter maternal and neonatal hospital durations post-delivery. These results demonstrate that the health outcome advantages associated with IOL are gained without incurring uniformly greater health care resource use.

Trial Registration:

ClinicalTrials.gov number NCT01990612

Condensation:

Women randomized to IOL compared to expectant management had significantly fewer antepartum visits, tests, and treatments, and shorter maternal and neonatal hospital durations post-delivery.

A randomized trial in which more than 6000 low-risk nulliparous women were randomized to planned elective labor induction at 39 weeks or expectant management, demonstrated that labor induction resulted in a lower frequency of cesarean delivery, hypertensive disorders of pregnancy, and neonatal respiratory morbidity.¹ After the publication of these results, the American College of Obstetricians and Gynecologists (ACOG) and the Society for Maternal-Fetal Medicine (SMFM) jointly stated that it was “reasonable for obstetricians and health-care facilities to offer elective induction of labor to low-risk nulliparous women at 39 weeks gestation.”² They also noted that additional information was needed to better understand the relationship between labor induction and health care resource utilization.^2,3

Indeed, several studies have suggested that elective induction of labor is associated with increased use of multiple obstetric interventions, such as regional analgesia, as well with higher costs.^4,5 Seyb et al., for example, estimated that elective labor induction increased costs by 17%.⁴ Other investigations have had similar conclusions.^6–8 Their studies, however, compared women undergoing labor induction with those in spontaneous labor. Such a comparison does not provide insight into the actual health-care resource consequences of labor induction, because the clinical alternative to labor induction is not spontaneous labor, but is expectant management. Additionally, such studies only accounted for differences in resource utilization and costs during the admission for delivery, and did not assess differences associated with the antepartum or postpartum period after discharge.

The purpose of this analysis was to determine whether planned elective labor induction at 39 weeks among low-risk nulliparous women, compared with expectant management, was associated with differences in health care resource utilization from the time of randomization through 8 weeks postpartum.

Methods

The ARRIVE trial was undertaken at 41 hospitals participating in the Maternal–Fetal Medicine Units Network of the Eunice Kennedy Shriver National Institute of Child Health and Human Development. Full details of the methods of this trial have been described previously.¹ In brief, reliably-dated low-risk nulliparous women with vertex, singleton gestations were eligible for the study. Those who consented to participation were randomized between 38 weeks 0 days and 38 weeks 6 days to either planned induction of labor at 39 weeks 0 days to 39 weeks 4 days or to expectant management. Those in the latter group were asked to forego non-medically indicated delivery before 40 weeks 5 days but to be delivered no later than 42 weeks 2 days. There were no other trial-specific requirements for antepartum, intrapartum, or postpartum care. Prior to initiation, the study was approved by the institutional review board at each hospital.

Following randomization, participants were followed up – both by review of medical records as well as by interviews performed by research personnel – to determine whether, prior to their admission for delivery, they had any outpatient visits or inpatient admissions. The former were classified according to whether they were routine visits for prenatal care or whether they were unscheduled visits other than for routine prenatal care. Unscheduled outpatient visits were further categorized according to their location of occurrence (i.e., ambulatory clinic, urgent care center, or hospital-based setting, such as an emergency department or obstetrical triage). When an inpatient hospitalization occurred, the total number of days was calculated as the duration between the day of admission and the day of discharge. Each outpatient visit or inpatient admission was characterized by the evaluations or interventions (e.g., imaging studies, laboratory tests, medications given) that were performed.

During the inpatient admission for delivery, the duration of time on the labor and delivery unit was determined as the difference between the time of admission to and the time of delivery. The duration of the postpartum stay was calculated as the difference in number of days between the day of delivery and the day of discharge. For both the woman and her newborn, interventions that occurred during the inpatient delivery admission were abstracted from the medical record. For the woman, these included interventions such as cervical ripening agents, oxytocin infusions, epidural use, and magnesium sulfate seizure prophylaxis, while for the neonate they included interventions such as phototherapy for hyperbilirubinemia and respiratory support.

Women were interviewed between 4 and 8 weeks after hospital discharge, at which time they were asked about any outpatient visits or inpatient admissions that they or their child had experienced, as well as any interventions (e.g., maternal or neonatal antibiotics) or tests (e.g., X-ray, complete blood count) they underwent. This interview was augmented for all women by medical record abstraction of documented health care through 8 weeks after discharge.

Analyses were performed according to the intention-to-treat principle. Resource utilization after randomization was categorized a priori according to when resources were used: antepartum, delivery admission, or from discharge through 8 weeks postpartum. Descriptive statistics were used to summarize data for women according to their randomized group assignment. We compared continuous variables using the Wilcoxon rank sum test and categorical variables using chi-square or Fisher’s exact tests, as appropriate. Measures of effect were estimated with relative risks and 95% confidence intervals were provided. For each type of resource that was significantly different between groups, the difference in the frequency of use per 1000 women was calculated.

No imputation for missing values was performed. A p-value < 0.05 was used to define statistical significance, and all tests were two tailed. Analyses were performed with SAS version 9.4 (SAS Institute Inc., Cary, NC).

Results

Of 6106 women who were randomized, 6096 (99.8%) had outcome data available; 3059 were assigned to planned labor induction and 3037 to planned expectant management. The study population’s characteristics, which were balanced between groups, are presented in Table 1.

Table 1:

Characteristics of the study population

	Induction of labor N = 3059	Expectant management N = 3037
Age (years)	24.6 ± 5.0	24.5 ± 5.2
BMI (kg/m2)	31.6 ± 6.2	31.6 ± 6.1
Race/ethnicity
Non-Hispanic black	23.1	23.0
Non-Hispanic white	43.4	44.7
Hispanic	28.3	26.5
Asian	2.8	3.5
Other, unknown, or more than one race	2.4	2.4
Smoked during pregnancy	7.3	8.0
Modified Bishop score < 5	62.7	64.3

Data presented as mean ± standard deviation or %.

Number of missing values: BMI (n=30), modified Bishop score (n=2).

In the antepartum period after randomization, women undergoing planned labor induction were less likely to have visits for routine prenatal care, as well as unscheduled outpatient visits and inpatient admissions (Table 2). Similarly, as illustrated in Table 3, they were less likely to undergo fetal assessments (such as non-stress tests, biophysical profiles, and different sonographic assessments), laboratory testing (such as complete blood counts or metabolic assessment), and multiple different types of treatments (such as intravenous hydration and antibiotics).

Table 2:

Proportion (%) of women with at least one antepartum encounter with the health care system, stratified by randomized group assignment

	Induction of labor N = 3059	Expectant management N = 3037	P	RR (95% CI)
Ambulatory
Office visit, routine prenatal care	32.4	68.4	<0.001	0.47 (0.45–0.50)
Office visit, unanticipated	0.5	2.6	<0.001	0.20 (0.12–0.34)
Urgent care/emergency department/ obstetric triage visit	16.2	44.3	<0.001	0.37 (0.33–0.40)
Inpatient
Hospital admission	0.8	2.2	<0.001	0.39 (0.25–0.61)

Data presented as %.

RR, relative risk; CI, confidence interval

Table 3:

Proportion (%) of women with at least one antepartum test or intervention, stratified by randomized group assignment

	Induction of labor N = 3059	Expectant management N = 3037	P	RR (95% CI)
Fetal surveillance and imaging
Non-stress tests	10.6	31.9	<0.001	0.33 (0.30–0.37)
Contraction stress test	0.2	0.3	0.48	0.70 (0.26–1.82)
Modified biophysical profile	0.3	1.1	<0.001	0.26 (0.13–0.55)
Biophysical profile	1.1	4.0	<0.001	0.27 (0.18–0.39)
Umbilical artery Doppler	0.03	0.3	0.02	0.12 (0.02–0.99)
Fetal position	2.7	6.5	<0.001	0.41 (0.32–0.53)
Amniotic fluid index	1.4	6.3	<0.001	0.22 (0.16–0.31)
Fetal growth	1.3	4.3	<0.001	0.31 (0.22–0.44)
Laboratory testing
Complete blood count	2.7	5.9	<0.001	0.45 (0.35–0.59)
Metabolic panel	0.6	2.0	<0.001	0.29 (0.17–0.49)
Preeclampsia panel*	1.0	3.2	<0.001	0.30 (0.20–0.46)
Other†	1.7	3.0	0.001	0.57 (0.41–0.80)
Urinalysis	1.7	6.2	<0.001	0.27 (0.20–0.36)
Treatment
Analgesic‡	0.5	2.3	0.001	0.21 (0.12–0.37)
Intravenous hydration	0.5	1.8	<0.001	0.27 (0.15–0.48)
Antibiotic	0.3	0.7	0.02	0.38 (0.17–0.85)
Other medication§	0.9	2.4	<0.001	0.36 (0.23–0.56)

Data presented as %.

RR, relative risk; CI, confidence interval

^*A complete blood count and chemistry panel focused upon preeclampsia (e.g., creatinine, liver function tests).

^†Includes blood tests (e.g., such as amylase and lipase) and cultures from urine, blood, and skin.

^‡Includes medications such as acetaminophen, butalbital, and opioids.

^§Includes medications such as anti-emetics, anti-histamines, and anti-virals.

During the delivery admission, women randomized to planned induction of labor spent a longer duration in labor and delivery, although they were more likely to have shorter inpatient stays after delivery (Table 4). Women in the planned induction of labor group more frequently received cervical ripening agents, oxytocin infusions, and intrauterine pressure catheters. Conversely, they received fewer magnesium sulfate infusions for seizure prophylaxis and parenteral antibiotics. Their newborns were more likely to have shorter inpatient admissions, and less frequently received continuous positive-airway pressure or high-flow oxygen.

Table 4:

Delivery admission resource utilization stratified by randomized group assignment

	Induction of labor N = 3059	Expectant management N = 3037	p	RR (95% CI)
Maternal
Labor and delivery duration (days)	0.83 (0.53, 1.2)	0.57 (0.37, 0.85)	<0.001	--------
Cervical ripening	62.8	28.7	<0.001	2.19 (2.06–2.33)
Oxytocin infusion	84.5	73.3	<0.001	1.15 (1.12–1.18)
Intrauterine pressure catheter	41.8	36.6	<0.001	1.14 (1.07–1.21)
Fetal scalp electrode	26.3	25.2	0.36	1.04 (0.96–1.13)
Magnesium sulfate infusion	1.9	2.9	0.02	0.67 (0.48–0.92)
Epidural analgesia	93.6	93.4	0.76	1.00 (0.99–1.02)
Antibiotic infusion	42.7	45.8	0.02	0.93 (0.88–0.99)
Uterotonic medication	10.7	10.0	0.38	1.07 (0.92–1.24)
Red blood cell transfusion	1.7	1.7	1.00	0.99 (0.68–1.45)
Fresh frozen plasma	0.2	0.2	1.00	0.99 (0.32–3.07)
Cryoprecipitate	0.1	0.1	0.69	0.66 (0.11–3.96)
Uterine balloon for tamponade	0.5	0.5	0.86	0.93 (0.46–1.88)
LOS post-delivery > 2 days	17.8	20.9	0.002	0.85 (0.77–0.94)
Neonatal
CPAP/High-flow oxygen	2.7	3.9	0.02	0.71 (0.54–0.94)
US/CT/MRI head imaging	1.0	1.1	0.98	0.99 (0.61–1.62)
Phototherapy	4.7	4.7	0.91	1.01 (0.81–1.27)
Exchange transfusion	0.2	0.2	0.75	0.83 (0.25–2.71)
Head or whole body cooling	0.3	0.4	0.36	0.66 (0.27–1.62)
LOS > 2 days	23.1	26.9	<0.001	0.86 (0.79–0.94)

Data presented as median (Q1, Q3) or %.

RR = relative risk; CI, confidence interval; LOS, length of stay

Q1, first quartile; Q3, third quartile; CPAP, continuous positive airway pressure; US, ultrasound; CT, computed tomography; MRI, magnetic resonance imaging.

Number of missing values: labor and delivery duration (n=2), neonatal length of stay (n=5).

The time to the postpartum follow up interview was similar in both groups (median 5.9 vs. 6.0 weeks, p = 0.76). Data regarding resource utilization after discharge from the hospital and through the postpartum are presented in Table 5. There was no difference between the two groups of women in the frequency of unscheduled ambulatory visits or inpatient admissions. No test or treatment differed among women from either group. Neonates born to women in the two groups also had a similar frequency of outpatient visits in acute care settings (i.e., urgent care or emergency departments) or inpatient admissions. Conversely, those whose mothers had been assigned to the planned induction of labor group were more likely to have ambulatory office visits other than for routine neonatal care. The only difference in testing or treatment for neonates was with regard to serum bilirubin, which was evaluated more frequently among those born to women assigned to labor induction (Table 6).

Table 5:

Proportion (%) of women and neonates, post-discharge from the delivery admission, with at least one encounter with the health care system, stratified by randomized group assignment

	Induction of labor N = 3059	Expectant management N = 3037	P	RR (95% CI)
Maternal
Ambulatory
Office visit, unanticipated	5.8	6.2	0.48	0.93 (0.76–1.13)
Urgent care visit	1.3	1.3	1.00	1.02 (0.66–1.58)
Emergency department visit	6.6	6.8	0.76	0.97 (0.80–1.17)
Inpatient
Hospital admission	2.3	2.4	0.67	0.93 (0.67–1.28)
Neonatal
Ambulatory
Office visit, unanticipated	15.3	13.5	0.04	1.14 (1.01–1.29)
Urgent care visit	2.0	1.7	0.57	1.12 (0.78–1.62)
Emergency department visit	8.5	8.7	0.82	0.98 (0.83–1.15)
Inpatient
Hospital admission	3.1	2.6	0.22	1.21 (0.90–1.62)

Data presented as %.

RR, relative risk; CI, confidence interval

Table 6:

Proportion (%) of women and neonates, post-discharge from the delivery admission, with at least test or treatment, stratified by randomized group assignment

	Induction of labor N = 3059	Expectant management N = 3037	P	RR (95% CI)
Maternal
Imaging
X-ray	1.0	1.1	1.00	0.99 (0.61–1.62)
Lower extremity sonogram	0.3	0.3	1.00	0.99 (0.37–2.64)
Computed tomography	1.1	1.0	0.61	1.16 (0.71–1.91)
Magnetic resonance imaging	0.2	0.1	1.00	1.24 (0.33–4.62)
Laboratory testing
Complete blood count	3.6	4.2	0.26	0.86 (0.67–1.10)
Metabolic panel	2.9	3.0	0.76	0.95 (0.71–1.27)
Other laboratory test*	3.4	3.6	0.68	0.94 (0.72–1.22)
Treatment
Intravenous hydration	1.9	1.9	1.00	1.01 (0.70–1.45)
Antibiotics	5.0	5.7	0.21	0.87 (0.71–1.08)
Blood transfusion	0.1	0.2	0.11	0.28 (0.06–1.36)
Wound opened	0.4	0.7	0.12	0.57 (0.28–1.15)
Curettage	0.1	0.3	0.09	0.33 (0.09–1.22)
Laparotomy	0.1	0.1	1.00	0.99 (0.20–4.92)
Neonatal
Imaging
X-ray	1.5	1.2	0.32	1.26 (0.82–1.93)
Laboratory testing
Complete blood count	2.3	2.2	0.86	1.04 (0.74–1.45)
Metabolic panel	1.8	1.9	0.64	0.91 (0.63–1.31)
Serum bilirubin	3.5	1.8	<0.001	1.91 (1.39–2.63)
Blood culture	0.9	1.1	0.44	0.81 (0.49–1.35)
Other laboratory test†	1.7	1.7	0.92	0.97 (0.67–1.42)
Treatments
Antibiotic	2.0	2.1	0.93	0.98 (0.69–1.38)
Other medication‡	4.9	4.5	0.47	1.09 (0.87–1.37)

Data presented as %.

RR, relative risk; CI, confidence interval

^*Includes blood tests (e.g., such as amylase and lipase) and cultures from urine, blood, and skin.

^†Includes blood tests (e.g., reticulocyte count) and cultures from urine, blood, cerebrospinal fluid, and skin.

^‡Includes medications such as albuterol, H2-antogonists, and steroid cream.

Absolute differences in the tests and treatments that differed between groups are presented in Table 7. In the antepartum period after randomization, those in the planned induction group (per 1000 women) had 1206 fewer ambulatory visits, 13 fewer inpatient days, and 698 fewer surveillance, imaging, blood and urine tests. They also were exposed to 59 fewer treatments. During the delivery admission, those in the planned induction of labor group (per 1000 women) had 250 additional days on labor and delivery, and received 476 more ripening agents, 112 more oxytocin infusions and 52 more intrauterine pressure catheters. Conversely, they received 10 and 31 fewer magnesium sulfate and antibiotic infusions, respectively, and they and their newborns cumulatively spent 158 fewer days in the hospital after delivery. Neonates born to women in the planned induction group visited their pediatricians’ offices 73 more times and had 49 more tests for serum bilirubin. The large majority of these bilirubin tests (81.7%) occurred less than three days after delivery.

Table 7:

Absolute differences per 1000 women in types of resources that significantly differed between groups

	Induction of labor	Expectant management	Difference
Antepartum
Ambulatory visits	569	1775	−1206
Office visit, routine prenatal care	376	1127	−751
Office visit, unanticipated	5	29	−24
Urgent care/emergency department/ obstetric triage visit	188	619	−432
Inpatient days	9	22	−13
Tests prior to delivery admission	294	992	−698
Non-stress test	124	459	−334
Sonogram*	76	258	−183
Laboratory test†	77	195	−118
Urinalysis	17	80	−63
Treatments	23	82	−59
Analgesic	5	27	−22
IV hydration	5	19	−14
Antibiotic	3	7	−4
Other medication‡	10	30	−20
Delivery admission
Labor and delivery (patient days)	908	657	250
Ripening agent	846	370	476
Balloon catheter	404	186	218
Laminaria	1	0	1
Cervidil	62	23	39
PGE1 or gel	378	161	218
Oxytocin infusion	845	733	112
Intrauterine pressure catheter	418	366	52
Magnesium sulfate infusion	19	29	−10
Antibiotic infusion	427	458	−31
Maternal postpartum LOS (days)	2136	2181	−45
Postpartum unit	2135	2176	−41
Intensive care unit	1	5	−4
CPAP/High-flow oxygen	27	39	−11
Neonatal LOS (days)	2373	2485	−113
Well-baby unit	1911	1935	−24
Intermediate or ICU	462	551	−89
Post-discharge through 4–8 weeks postpartum
Office, visit unanticipated	288	215	73
Serum bilirubin	93	43	49

Difference = Induction of labor – Expectant Management.

ED, emergency department; PGE1, prostaglandin E1; LOS, length of stay; HFO2, high-flow oxygen; CPAP, continuous positive airway pressure; ICU, intensive care unit.

^*Includes all sonographic exams (i.e., biophysical profiles, modified biophysical profiles, umbilical artery Dopplers, amniotic fluid volume, fetal position, and fetal growth).

^†Includes all blood tests and cultures.

^‡Includes medications such as anti-emetics, anti-histamines, and anti-depressants Number of missing values: labor and delivery duration (n=2), neonatal length of stay (n=5).

Discussion

Principal Findings

We have demonstrated that women randomized to planned labor induction at 39 weeks differed significantly from those randomized to planned expectant management in terms of health care resource utilization. Specifically, women assigned to planned induction used significantly fewer resources prior to delivery. During the delivery admission, the groups were more balanced, with some resources used more frequently by those in the induction group and other resources by those in the expectant management group. Postpartum resources were largely similar, with the only difference being that neonates born to women in the induction group were slightly more likely to have an outpatient pediatric visit or an outpatient serum bilirubin tested. Because there was no difference between groups in the frequency of or treatments for hyperbilirubinemia,’ and the large majority of outpatient bilirubin tests were performed so shortly after delivery, this difference seems largely to be due to fact that infants in the induction group – who were discharged from the hospital sooner than their expectantly managed counterparts – were being evaluated as outpatients rather than inpatients for this level.

Results

These results are in contrast to results of prior studies, which concluded– based on analyses in which women undergoing labor induction were compared to those in spontaneous labor – that elective induction of labor was associated with greater use of many different types of resources. For example, van Gemund et al. found that women who underwent elective induction were more likely to receive epidural analgesia and have longer hospital stays.³ Cammu et al. also documented more frequent epidural use, as well as a greater likelihood of NICU admission after elective induction.⁸ Seyb et al.⁵ suggested that more resources were used not just in labor and delivery but also during the postpartum stay. An increase in costs among women undergoing induction, similar in magnitude to those documented by Seyb et al, was found in other studies in the United States as well as in Canada.^6,7,9

Research and Clinical implications

Such studies, however, may give very different results than those performed with the clinically-relevant comparator of expectant management.^10,11 Indeed, our analysis refutes the belief that elective induction of labor is uniformly more resource intensive than expectant management. This is particularly true during a period not examined in these prior studies – the antepartum period – during which expectant management is significantly more resource intensive. Even during the delivery admission, which was the focus of prior studies, the results of our study do not substantiate prior findings. For example, in our study, epidural analgesia was similar in both groups, and postpartum duration was actually shorter for both women and children in the induction group.

Strengths and Limitations

There are multiple strengths of this study, most prominently that the results reflect the consequences of the actual clinical decision of labor induction at 39 weeks vs. expectant management among low-risk nulliparous women, and are derived from a randomized trial, minimizing both confounding bias and the possibility that the relationships seen are merely associative. In addition, detailed assessment of specific resources, defined by an a priori plan, from the antepartum through 8 weeks postpartum add novel information and fill an important knowledge gap.^2,3

A limitation of this analysis is that we were unable to collect actual costs due to several factors, including that there is no universally-understood cost in the United States’ healthcare system for the different resources, and different hospitals would not share their own actual costs for services, such as time on labor and delivery, which were considered proprietary information. Cost determination, therefore, would have required the use of multiple assumptions and estimations, and not be actual data from the trial itself. Given these constraints, we believe it is ultimately preferable to primarily present and focus upon the novel, detailed, and reliable information about resource utilization and avoid the inherent weaknesses and imprecision of modeling exercises. Nevertheless, even if one wished to estimate the tradeoffs in the most substantive costs associated with the resources utilized, the data do not support the concept that labor induction will be markedly more costly than expectant management. While the additional costs of labor induction are largely related to additional time on the labor unit, the women who undergo expectant management have more ambulatory visits, inpatient hospitalizations, antepartum surveillance tests, hypertensive disorders of pregnancy, and cesarean deliveries; their neonates require more respiratory support and have a greater number of days in an intermediate or intensive care nursery. Even a broad approximation suggests that the net costs are similar – and may even be less for labor induction (supplementary table 1). Such a conclusion is consistent with the results from analyses of the three randomized trials performed outside of the United States that compared costs of induction versus expectant management in different populations –- women at 41 weeks, women with hypertensive disorders of pregnancy, and women over 35 years of age at 39 weeks.^12–14 In the first two of these analyses, labor induction was the less expensive option, and in the third it was similar in cost. Driving reductions in costs in these studies were reductions in antepartum expenses; this finding mirrors the significant reduction in resource use seen during the antepartum period in our study.

Conclusions

In the ARRIVE trial, low-risk women randomized to planned labor induction at 39 weeks had better maternal and neonatal outcomes.¹ The present analysis demonstrates that the health outcome advantages associated with IOL are gained without incurring uniformly greater health care resource use. Further analyses – particularly within individual hospitals or healthcare networks in which actual costs can be analyzed – can provide further perspective on the health-system implications of labor induction among low-risk nulliparous women at 39 weeks of gestation.

AJOG at a Glance.

1. Why was this study conducted?

   • Comparative resource utilization for women undergoing labor induction versus expectant management has not been well understood given the reliance on observational study designs and focus only on intrapartum care.
   • This study is able to use data from a randomized trial of labor induction at 39 weeks versus expectant management to delineate differences in resource use during the antepartum, intrapartum, and postpartum periods.
2. What are the key findings?

   • Women who undergo labor induction have fewer antepartum ambulatory visits and hospitalizations, undergo many fewer tests and treatments, and have a shorter postpartum inpatient duration.
   • Women who undergo labor induction have longer duration on labor and delivery
3. What does this study add to what is already known?

   • This study provides novel information regarding differential resource use based on 39-week labor induction versus expectant management.

Supplementary Table 1:

Net cost differences (2018 US dollars) for labor induction versus expectant management

	Difference*	Unit cost	Total cost*
AP ambulatory office visits	−775	136.3915	−105,702.25
AP triage/urgent care/ED visits	−432	215.4715	−93,083.04
AP inpatient days	−13	2326.7616	−30,247.85
AP sonograms	−183	76.1417	−13,933.65
AP non-stress tests	−334	35.6917	−11,919.59
HDP	−501	3650.1815	−182,509.00
Days on labor and delivery	250	2233.9612	558,490.81
Cesarean delivery†	−361	4682.8115	−168,581.16
Neonatal intermediate or intensive care nursery days	−89	1077.5316	−95,900.36
		Net	−143,386.08
		Net per woman	−143.39

Difference = Induction of labor − Expectant Management.

AP, antepartum; ED, emergency department; HDP, hypertensive disorders of pregnancy; CPAP, continuous positive airway pressure

All costs adjusted to 2018 US dollars^18,19

^*Per 1000 women

^†Excess cost of cesarean delivery compared to vaginal delivery